Lamb Wave Propagation in Complex Geometries - A Minimal Model Approach

Kurzfassung

Maintenance of aircraft structures is time consuming and thus expensive. Several factors contribute to this, for instance a lack of automation. Furthermore, inspection methods for emerging materials, like carbon fiber reinforces polymers (CFRP), are not yet mature. Uncertainties regarding the long time behavior of damaged and undamaged CFRP lead to high safety margins in structural design. A system for online inspection of large areas could be a solution to this. Different methods for structural health monitoring (SHM) have been explored in the last decades, but active systems with ultrasonic waves are the most promising ones. Interactions in the aircraft structure result in complex wave fields and damage induced signal changes are hard to identify. Simulations could help understand wave propagation, but neither commercial programs nor any known algorithm currently in development can handle these dynamic processes for large areas in acceptable times. All these methods use detailed models to consider all relevant influences. Contrary to this, a minimal model is proposed to simulate Lamb wave propagation in complex geometries. Model reduction is achieved by approximation of repeatedly used structural elements, like stiffeners. The model consists of areas with individual material properties and specific boundary conditions at their border. Interaction behavior at these inhomogeneities is calculated beforehand and summarized with few characteristic values. The most important parameters are transmission, reflection and the conversion of different wave modes. Ray tracing is used to find the fastest paths in the structure between two arbitrary points, i.e. actuator and sensor. The gained information and material specific values are used to calculate the time signal. This approximate method could be used is SHM systems to adapt reference signals to changing environmental conditions, as intelligent reference signal or to optimize actuator sensor networks.